CN112130154A - Self-adaptive K-means outlier de-constraint optimization method for fusion grid LOF - Google Patents
Self-adaptive K-means outlier de-constraint optimization method for fusion grid LOF Download PDFInfo
- Publication number
- CN112130154A CN112130154A CN202010850065.0A CN202010850065A CN112130154A CN 112130154 A CN112130154 A CN 112130154A CN 202010850065 A CN202010850065 A CN 202010850065A CN 112130154 A CN112130154 A CN 112130154A
- Authority
- CN
- China
- Prior art keywords
- outlier
- target
- positioning
- data
- algorithm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 230000004927 fusion Effects 0.000 title claims abstract description 37
- 238000005457 optimization Methods 0.000 title claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 18
- 238000012216 screening Methods 0.000 claims abstract description 10
- 230000003044 adaptive effect Effects 0.000 claims description 7
- 238000002592 echocardiography Methods 0.000 claims description 6
- 238000013450 outlier detection Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 4
- 230000008030 elimination Effects 0.000 abstract 2
- 238000003379 elimination reaction Methods 0.000 abstract 2
- 238000005259 measurement Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 6
- 238000013461 design Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 235000021178 picnic Nutrition 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S15/00—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems
- G01S15/003—Bistatic sonar systems; Multistatic sonar systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S15/00—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems
- G01S15/02—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems using reflection of acoustic waves
- G01S15/06—Systems determining the position data of a target
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
Abstract
The invention discloses a self-adaptive K-means outlier de-constraint optimization method for fusing grid LOF. The invention relates to the technical field of underwater target positioning, which acquires target positioning coordinate data through a sensor, preprocesses the target positioning coordinate data, eliminates error data to obtain target positioning data, establishes a bistatic sonar system positioning model, establishes a self-adaptive K-means outlier algorithm model of a fusion grid LOF (low order filtered) and a constraint elimination model, performs constraint elimination processing on a discrete point set, performs outlier judgment and outputs an outlier sensor point set; and returning the data set of the outlier sensor node, screening an effective sensor point set, and positioning a target according to the effective sensor point set to improve the target positioning precision. The invention can accurately obtain the sensor nodes with larger errors, eliminate the data measured by the sensor nodes, effectively compress the data volume and improve the positioning performance.
Description
Technical Field
The invention relates to the technical field of underwater target positioning, in particular to a self-adaptive K-means outlier de-constrained optimization method fusing a grid LOF.
Background
In recent years, along with the weakening of radiation signals of 'quiet' nuclear submarines and the randomness and complexity of space-time change of the internal environment of the ocean, the requirement on the positioning accuracy of an underwater wireless sensor network is gradually increased. In an underwater wireless sensor network, the positioning accuracy of a target is often limited by the magnitude of an estimation error. In general, error generation can be divided into subjective and objective aspects. Errors in subjective aspects are mainly caused by inappropriate occupation ratio of anchor nodes in a sensing network and imperfection (large calculation amount, low calculation speed and the like) of a positioning algorithm; the objective error is mainly caused by the influence of factors such as the nonuniformity of node distribution, the failure of partial nodes, the complexity of environment or equipment obstacle. Therefore, algorithm optimization from the above two error aspects is a reliable direction to improve positioning accuracy.
Currently, research on how to effectively reduce objective errors is more than subjective errors. The following two directions can be roughly divided: firstly, different algorithms are used for solving a target positioning equation, such as a least square method with constraint, maximum likelihood estimation and the like; and secondly, performing weighted fusion on the measured data according to the weight. Although the two research directions for reducing objective errors can obviously improve the positioning precision, the measurement errors caused by partial node failure cannot be effectively removed. Accordingly, some scholars propose new solutions. And (3) the friend and the like perform data quality analysis processing on all positioning points according to the distance square sum (density), and finally, the data center with high density is used as a target estimation position, so that the estimation error is reduced. The probability operation is introduced into a target positioning algorithm by the people of picnic, Tan Kun, Ponfaphenant and the like, and the maximum peak point corresponding to the probability density function is finally used as the estimated position of the target by solving the probability function, so that the random measurement error is reduced.
Research on reducing subjective errors is also continuously developing, and particularly in the aspect of improvement of positioning algorithms, certain achievements have been achieved at present, that is, different improved positioning algorithms are proposed, such as a Zigbee positioning algorithm, an RSSI-based ranging positioning algorithm, a kalman filter positioning algorithm, and the like. But the research on reducing the influence of error anchor nodes (sensor nodes) and improving the accuracy of observed values is yet to be perfected.
In an underwater target sensing and positioning network, positioning of a target needs to be completed by cooperation of a plurality of anchor nodes, and the positioning can be completed according to a measured distance, coordinate calculation, azimuth estimation or other algorithms and the like. Generally, one anchor node corresponds to one sensor node. In addition, the measured data is derived from the sensor nodes, so the estimation error of the positioning is mainly derived from the measurement error of the sensor nodes, and the error is influenced by factors such as marine environment, equipment and the like.
Disclosure of Invention
The invention provides a self-adaptive K-means outlier de-constraint optimization method for fusing grid LOF (loss of tolerance), which is used for compressing effective data and improving positioning precision by screening data with larger error deviation, namely outlier clusters, and provides the following technical scheme:
a self-adaptive K-means outlier de-constrained optimization method for fusion grid LOF comprises the following steps:
step 1: acquiring coordinate data of target positioning through a sensor, preprocessing the coordinate data of the target positioning, and removing error data to obtain target positioning data;
step 2: establishing a bistatic sonar system positioning model, inputting the target positioning data into the bistatic sonar system positioning model, and outputting an estimated value of a target position solution under an ideal condition;
and step 3: establishing a self-adaptive K-means outlier algorithm model of the fusion grid LOF, and inputting an estimated value of a target position solution under an ideal condition into the self-adaptive K-means outlier algorithm model of the fusion grid LOF to obtain an outlier set;
and 4, step 4: establishing a constraint removing model, carrying out constraint removing processing on the outlier set, carrying out outlier judgment, and outputting the outlier sensor point set;
and 5: and returning the data set of the outlier sensor node, screening an effective sensor point set, and positioning a target according to the effective sensor point set to improve the target positioning precision.
Preferably, in the step 1, a fusion grid LOF algorithm is adopted to preprocess the coordinate data of the target positioning, to screen out outliers which are deviated from the center and isolated, and to remove error data to obtain the target positioning data.
Preferably, the step 2 specifically comprises:
step 2.1: establishing a bistatic sonar system positioning model, wherein the model comprises a transmitting end T end and a receiving end R end, the T end is responsible for transmitting sound waves, is deployed in a protected area and receives target echoes at the same time, and belongs to independent sonars; the R end does not emit sound waves and only receives target echoes, so that the concealment is high; and the S end estimates the position of the target, and determines a direction angle positioning equation according to a bistatic sonar positioning algorithm:
wherein, alpha is the beam pointing angle of the position of the target measured at the transmitting end, beta is the beam pointing angle of the receiving end, and x0，y0Estimating position coordinates, x, for the targetT，yTAs a transmitting end coordinate, yR，xRIs the coordinate of the receiving end;
inputting the target positioning data into an orientation angle positioning equation in a bistatic sonar system positioning model, outputting an estimated value of a target position solution under an ideal condition, and expressing the estimated value by the following formula:
preferably, the step 3 specifically comprises:
establishing a self-adaptive K-means outlier algorithm model of a fusion grid LOF, inputting an estimated value of a target position solution under an ideal condition into the self-adaptive K-means outlier algorithm model of the fusion grid LOF as input data, unifying the estimated value of the target position solution under the ideal condition in the aspects of format, unit and the like in a standardization processing process of data in the self-adaptive K-means outlier algorithm model of the fusion grid LOF, carrying out grid division of a limited unit on the input data by the grid LOF algorithm, independently operating LOF algorithm on each part, outputting isolated outliers through the LOF algorithm, and simultaneously outputting a data set which does not contain the outliers for carrying out outlier detection by the self-adaptive K-means algorithm;
the self-adaptive K-means outlier algorithm firstly sets an initial value K to be 2, the step length is 1, a plurality of different K values are obtained, K-means outlier detection is sequentially carried out on the K values by using a slope ratio method, the optimal K value is obtained, K data sets are obtained, and the data set with the least number of data is classified into an outlier set.
Preferably, the clustering number K in the K-means outlier algorithm is operated to be 1.
Preferably, the step 4 specifically includes:
step 4.1: establishing a constraint removing model, carrying out constraint removing processing on the outlier set, determining the corresponding relation between the sensor nodes and the outliers, and expressing a relation equation between the sensor nodes and the outliers through the following formula:
wherein S is an outlier, j is a sensor number, TiThe corresponding sensor nodes are arranged in the corresponding sensor nodes,
step 4.2: counting the times M of the sensor nodes with the outlier error according to the corresponding relation, calculating the influence weight and setting a reasonable judgment threshold value, and expressing the influence weight gamma by the following formula:
wherein M isiRepresenting the number of times of generating the outlier error of the ith node;
step 4.3: performing outlier judgment, performing compromise processing on a threshold value xi, wherein when xi is 0.5, the result of an outlier error is minimum, and the positioning accuracy of a target is highest;
when gamma is larger than or equal to xi, outputting an outlier sensor point set; otherwise, outputting the effective sensor point set.
Preferably, the sum of squares of the distances from each sensor node to the clustering center of the K-means algorithm is calculated, after the K-means algorithm is clustered, the sum of squares of the distances from each point in the region to the clustering center is minimum to serve as a standard, the number of times of cyclic calculation is counted, the number of times of exceeding the standard is the number of times of the sensor node being in outlier, and the number of times is recorded as M.
Preferably, the step 5 specifically comprises: the method comprises the steps of returning an outlier sensor node data set, carrying out subtraction operation on the outlier sensor node data set and an initial sensor node data set, screening out an effective sensor point set, inputting the effective sensor point set into a bistatic sonar system positioning model, and effectively positioning a target.
The invention has the following beneficial effects:
the invention can effectively simplify the data, reduce the calculation amount, accelerate the operation speed and improve the positioning precision. Realizing initial positioning through a bistatic sonar system to obtain a huge data set containing errors, and screening out a large amount of error positioning data, namely an outlier set, through a self-adaptive K-means outlier algorithm of a fusion grid LOF; and finally, combining a constraint removing algorithm, obtaining sensor nodes with larger errors accurately, eliminating data measured by the sensor nodes, effectively compressing data volume and improving positioning performance.
The bistatic sonar system model provided by the invention realizes the split arrangement of the receiving and transmitting equipment, so that the bistatic sonar system model has the working characteristics of active sonar and passive sonar at the same time, and has obvious potential superiority in the aspects of anti-stealth and anti-objection. The self-adaptive K-means outlier algorithm fused with the grid LOF can process huge data sets, has high operation speed, has low time and space complexity, moderate realization difficulty and better processing effect than that of a single outlier algorithm. The constraint removing algorithm model only needs to calculate the outlier sensor node set, so the realization process is simple, the processing speed is high, and the performance is relatively good.
Drawings
FIG. 1 is a flow chart of a method for adaptive K-means outlier de-constrained optimization with mesh LOF fusion;
FIG. 2 is a model diagram of a bistatic sonar positioning system;
FIG. 3 is a flow chart of the design of the adaptive K-means outlier algorithm for fusion mesh LOF;
FIG. 4 is a flow chart of the design of the de-constraining algorithm.
Detailed Description
The present invention will be described in detail with reference to specific examples.
The first embodiment is as follows:
as shown in fig. 1, the invention provides a self-adaptive K-means outlier de-constraint optimization method for fusion grid LOF, which specifically comprises the following steps:
a self-adaptive K-means outlier de-constrained optimization method for fusion grid LOF comprises the following steps:
step 1: acquiring coordinate data of target positioning through a sensor, preprocessing the coordinate data of the target positioning, and removing error data to obtain target positioning data;
in the step 1, a fusion grid LOF algorithm is adopted to preprocess coordinate data of target positioning, outliers which are deviated from a center and isolated are screened out, and error data are removed to obtain target positioning data.
Step 2: establishing a bistatic sonar system positioning model, inputting the target positioning data into the bistatic sonar system positioning model, and outputting an estimated value of a target position solution under an ideal condition;
the step 2 specifically comprises the following steps:
step 2.1: establishing a bistatic sonar system positioning model, wherein the model comprises a transmitting end T end and a receiving end R end, the T end is responsible for transmitting sound waves, is deployed in a protected area and receives target echoes at the same time, and belongs to independent sonars; the R end does not emit sound waves and only receives target echoes, so that the concealment is high; and the S end estimates the position of the target, and determines a direction angle positioning equation according to a bistatic sonar positioning algorithm:
wherein, alpha is the beam pointing angle of the position of the target measured at the transmitting end, beta is the beam pointing angle of the receiving end, and x0，y0Estimating position coordinates, x, for the targetT，yTAs a transmitting end coordinate, yR，xRIs the coordinate of the receiving end;
inputting the target positioning data into an orientation angle positioning equation in a bistatic sonar system positioning model, outputting an estimated value of a target position solution under an ideal condition, and expressing the estimated value by the following formula:
the above is only a target position solution obtained under ideal conditions, and errors caused by marine environments or actual measurement are not considered, so that it is necessary to study the influence of errors on the target position solution.
Defining an angle measurement error equation:
where α 'and β' represent measured values with noise, and d α and d β represent observed random errors. Since it is practically impossible to obtain ideal data, it is common to replace the ideal values with measured data values, i.e. α 'instead of α, β' instead of β. It is assumed that the measurement error of each angle is zero-mean white gaussian noise which is uncorrelated with each other, and the standard deviation of each angle is sum. The orientation angle positioning equation is differentiated to obtain:
wherein the content of the first and second substances,
the above equation is expressed as a matrix:
dP＝BdQ+dQs
wherein dP ═ d α d β]T、dQ＝[dx0 dy0]T、dQs＝[kα kβ]T、
Thus, the target position error vector can be solved:
dQ＝B-1[dP-dQs]
the position estimation of the target can be subjected to error analysis according to the error vector, and the target position estimation error is related to the angle measurement error and the station address measurement error. Due to the influence of factors such as equipment and the internal environment of the ocean, the error is difficult to reduce, and in order to improve the performance of the positioning effect, other algorithms are required to be matched for reducing the error.
In summary, the estimated value of the target position can be obtained through the bistatic sonar system positioning model. In addition, the coordinates of one target positioning point can be calculated by the direction angles measured by the two sensor nodes. In practice, a plurality of sensor nodes are generally arranged, and if n is assumed, the group target position coordinate data C can be obtained through the system positioning modeln2. The method comprises certain error data, needs to be screened, and retains effective position data to improve the positioning performance, namely, provides a self-adaptive K-means outlier algorithm of the fusion grid LOF.
And step 3: establishing a self-adaptive K-means outlier algorithm model of the fusion grid LOF, and inputting an estimated value of a target position solution under an ideal condition into the self-adaptive K-means outlier algorithm model of the fusion grid LOF to obtain an outlier set;
the step 3 specifically comprises the following steps:
establishing a self-adaptive K-means outlier algorithm model of a fusion grid LOF, inputting an estimated value of a target position solution under an ideal condition into the self-adaptive K-means outlier algorithm model of the fusion grid LOF as input data, unifying the estimated value of the target position solution under the ideal condition in the aspects of format, unit and the like in a standardization processing process of data in the self-adaptive K-means outlier algorithm model of the fusion grid LOF, carrying out grid division of a limited unit on the input data by the grid LOF algorithm, independently operating LOF algorithm on each part, outputting isolated outliers through the LOF algorithm, and simultaneously outputting a data set which does not contain the outliers for carrying out outlier detection by the self-adaptive K-means algorithm;
the self-adaptive K-means outlier algorithm firstly sets an initial value K to be 2, the step length is 1, a plurality of different K values are obtained, K-means outlier detection is sequentially carried out on the K values by using a slope ratio method, the optimal K value is obtained, K data sets are obtained, and the data set with the least number of data is classified into an outlier set.
The design flow chart of the self-adaptive K-means outlier algorithm of the fusion grid LOF is shown in FIG. 3. The input data stream is target coordinate data obtained through calculation of a bistatic sonar system positioning model, and the target coordinate data comprises a certain amount of error data. Because actual underwater target positioning generally adopts a plurality of sensor nodes to position a target, the value of the clustering number K in the operation K-means outlier algorithm is 1, and the exact number 1 is not contradictory to the dynamically changed K value in the self-adaptive K-means algorithm. Because the sensor node positions the same target, most coordinate information has aggregation and the distribution is relatively concentrated, even if the adaptive K-means algorithm takes the value of the K value to be more than 1, most data sets still present central aggregation distribution, and therefore the positioning of the single target is not influenced. But the k value is subjected to a dynamic value taking process, so that the whole data set is finely split, and the outliers can be conveniently screened out. And before the K-means outlier algorithm detection is carried out, the grid LOF algorithm is operated, so that the time and space complexity of the whole algorithm is reduced, and the whole performance of the algorithm is better.
And 4, step 4: establishing a constraint removing model, carrying out constraint removing processing on the outlier set, carrying out outlier judgment, and outputting the outlier sensor point set;
the flow chart of the design of the de-constraining algorithm is shown in FIG. 4. The method comprises the steps of firstly determining the corresponding relation between sensor nodes and outliers, counting the times of the sensor nodes with outlier errors according to the corresponding relation, calculating influence weights and setting a reasonable judgment threshold, regarding the sensor nodes larger than the threshold as the outliers, and reserving the sensor nodes smaller than the outliers.
The step 4 specifically comprises the following steps:
step 4.1: establishing a constraint removing model, carrying out constraint removing processing on the outlier set, determining the corresponding relation between the sensor nodes and the outliers, and expressing a relation equation between the sensor nodes and the outliers through the following formula:
wherein S is an outlier, j is a sensor number, TiThe corresponding sensor nodes are arranged in the corresponding sensor nodes,
step 4.2: counting the times M of the sensor nodes with the outlier error according to the corresponding relation, calculating the influence weight and setting a reasonable judgment threshold value, and expressing the influence weight gamma by the following formula:
wherein M isiRepresenting the number of times of generating the outlier error of the ith node;
step 4.3: performing outlier judgment, performing compromise processing on a threshold value xi, wherein when xi is 0.5, the result of an outlier error is minimum, and the positioning accuracy of a target is highest;
when gamma is larger than or equal to xi, outputting an outlier sensor point set; otherwise, outputting the effective sensor point set.
And calculating the square sum of the distances from each sensor node to the clustering center of the K-means algorithm, circularly calculating a plurality of times by taking the minimum square sum of the distances from each point in the region to the clustering center as a standard after the K-means algorithm is clustered, and counting the times exceeding the standard, namely the times of the sensor node clustering, and recording as M. E.g. 20 times (more than 10 times), counting the times exceeding the standard, namely the times of the sensor node being in the outlier
And 5: and returning the data set of the outlier sensor node, screening an effective sensor point set, and positioning a target according to the effective sensor point set to improve the target positioning precision.
The step 5 specifically comprises the following steps: the method comprises the steps of returning an outlier sensor node data set, carrying out subtraction operation on the outlier sensor node data set and an initial sensor node data set, screening out an effective sensor point set, inputting the effective sensor point set into a bistatic sonar system positioning model, and effectively positioning a target.
Description of the principle:
bistatic sonar system positioning model
The bistatic sonar system positioning model is the simplest and most basic configuration of a multi-base sonar system. The position of the unknown target point can be determined through the direction intersecting line or the distance intersecting line between the two sonars and the target position. In one plane, two unparallel straight lines are intersected necessarily, and accordingly, the intersection point of the direction intersection line or the distance intersection line formed between the two unparallel sonars and the target is used as the estimated position of the final target. In the present invention, the intersection of two directional intersecting lines (angular intersecting lines) is used as the estimated position of the target point. A data set containing target positioning coordinate information can be obtained by solving the direction angle positioning equation, and factors which may generate errors are analyzed and researched.
A bistatic sonar system positioning model is a common simple configuration model in a multi-base sonar system. Because transceiver separately places for bistatic sonar system possesses the characteristics and the performance of active and passive sonar working method simultaneously. The system model is shown in fig. 2.
Wherein the T end is the transmitting terminal, is responsible for the transmission sound wave, generally deploys in the protected area, also can receive the target echo simultaneously, belongs to independent sonar. R terminal isAnd the receiving end does not transmit sound waves and only receives the target echo, so the concealment is high. And the S end is a target estimation position and needs to be accurately positioned. r is1Denotes the distance between the transmitting end and the target, r2D represents the base length, which is the distance between the receiving end and the target, and in a bistatic system, the base length is defined as the distance between the transmitting end and the receiving end. Alpha is the beam pointing angle of the position of the target measured at the transmitting end, beta is the beam pointing angle of the receiving end, and theta is the separation angle, and defines the included angle of the connecting line between the transmitting end, the receiving end and the target when viewed from the target point.
Self-adaptive K-means outlier algorithm model of fusion grid LOF
The self-adaptive K-means outlier algorithm of the fusion grid LOF is essentially based on the combination of the local abnormal factor (LOF) algorithm of the grid and the self-adaptive K-means algorithm. Firstly, preprocessing initial data by utilizing a grid LOF algorithm, and screening out outliers which are deviated from a center and are isolated; and then, accurately detecting outliers of the remaining data sets by adopting a self-adaptive K-means algorithm, finally summarizing the outliers filtered by the two parts, and outputting an outlier set for subsequent constraint processing.
De-constrained algorithm model
The constraint-removing algorithm model is used for carrying out subsequent processing on the outlier output point set, and the outlier set is subjected to constraint removing to obtain an outlier sensor node (anchor node) set, so that data (with large errors) obtained by testing the outlier sensor node are removed, effective data are simplified, and positioning accuracy is improved. In the bistatic sonar system positioning model, one target positioning node corresponds to two sensor nodes (anchor nodes). Determining the corresponding relation between the outliers and the sensor nodes according to the principle, thereby counting the times of deviation of each sensor node from the center and the outliers, calculating the influence weight, setting a reasonable threshold value, regarding the sensor nodes larger than the threshold value as the outlier sensor nodes, and then removing the sensor nodes. The three parts can be circulated for multiple times to improve the performance of discriminating data with larger errors, so that effective data containing target positioning information is finally left.
The above description is only a preferred embodiment of the adaptive K-means outlier de-constraining optimization method for the fusion grid LOF, and the protection scope of the adaptive K-means outlier de-constraining optimization method for the fusion grid LOF is not limited to the above embodiments, and all technical solutions belonging to the idea belong to the protection scope of the present invention. It should be noted that modifications and variations which do not depart from the gist of the invention will be those skilled in the art to which the invention pertains and which are intended to be within the scope of the invention.
Claims (8)
1. A self-adaptive K-means outlier de-constrained optimization method for fusion grid LOF is characterized by comprising the following steps: the method comprises the following steps:
step 1: acquiring coordinate data of target positioning through a sensor, preprocessing the coordinate data of the target positioning, and removing error data to obtain target positioning data;
step 2: establishing a bistatic sonar system positioning model, inputting the target positioning data into the bistatic sonar system positioning model, and outputting an estimated value of a target position solution under an ideal condition;
and step 3: establishing a self-adaptive K-means outlier algorithm model of the fusion grid LOF, and inputting an estimated value of a target position solution under an ideal condition into the self-adaptive K-means outlier algorithm model of the fusion grid LOF to obtain an outlier set;
and 4, step 4: establishing a constraint removing model, carrying out constraint removing processing on the outlier set, carrying out outlier judgment, and outputting the outlier sensor point set;
and 5: and returning the data set of the outlier sensor node, screening an effective sensor point set, and positioning a target according to the effective sensor point set to improve the target positioning precision.
2. The method of claim 1, wherein the method comprises: in the step 1, a fusion grid LOF algorithm is adopted to preprocess coordinate data of target positioning, outliers which are deviated from a center and isolated are screened out, and error data are removed to obtain target positioning data.
3. The method of claim 1, wherein the method comprises: the step 2 specifically comprises the following steps:
step 2.1: establishing a bistatic sonar system positioning model, wherein the model comprises a transmitting end T end and a receiving end R end, the T end is responsible for transmitting sound waves, is deployed in a protected area and receives target echoes at the same time, and belongs to independent sonars; the R end does not emit sound waves and only receives target echoes, so that the concealment is high; and the S end estimates the position of the target, and determines a direction angle positioning equation according to a bistatic sonar positioning algorithm:
wherein, alpha is the beam pointing angle of the position of the target measured at the transmitting end, beta is the beam pointing angle of the receiving end, and x0，y0Estimating position coordinates, x, for the targetT，yTAs a transmitting end coordinate, yR，xRIs the coordinate of the receiving end;
inputting the target positioning data into an orientation angle positioning equation in a bistatic sonar system positioning model, outputting an estimated value of a target position solution under an ideal condition, and expressing the estimated value by the following formula:
4. the method of claim 1, wherein the method comprises: the step 3 specifically comprises the following steps:
establishing a self-adaptive K-means outlier algorithm model of a fusion grid LOF, inputting an estimated value of a target position solution under an ideal condition into the self-adaptive K-means outlier algorithm model of the fusion grid LOF as input data, unifying the estimated value of the target position solution under the ideal condition in the aspects of format, unit and the like in a standardization processing process of data in the self-adaptive K-means outlier algorithm model of the fusion grid LOF, carrying out grid division of a limited unit on the input data by the grid LOF algorithm, independently operating LOF algorithm on each part, outputting isolated outliers through the LOF algorithm, and simultaneously outputting a data set which does not contain the outliers for carrying out outlier detection by the self-adaptive K-means algorithm;
the self-adaptive K-means outlier algorithm firstly sets an initial value K to be 2, the step length is 1, a plurality of different K values are obtained, K-means outlier detection is sequentially carried out on the K values by using a slope ratio method, the optimal K value is obtained, K data sets are obtained, and the data set with the least number of data is classified into an outlier set.
5. The method of claim 4, wherein the adaptive K-means outlier de-constrained optimization method for fusion mesh LOF comprises: and operating a clustering number K in a K-means outlier algorithm to take the value of 1.
6. The method of claim 1, wherein the method comprises: the step 4 specifically comprises the following steps:
step 4.1: establishing a constraint removing model, carrying out constraint removing processing on the outlier set, determining the corresponding relation between the sensor nodes and the outliers, and expressing a relation equation between the sensor nodes and the outliers through the following formula:
wherein S is an outlier, j is a sensor number, TiThe corresponding sensor nodes are arranged in the corresponding sensor nodes,
step 4.2: counting the times M of the sensor nodes with the outlier error according to the corresponding relation, calculating the influence weight and setting a reasonable judgment threshold value, and expressing the influence weight gamma by the following formula:
wherein M isiRepresenting the number of times of generating the outlier error of the ith node;
step 4.3: performing outlier judgment, performing compromise processing on a threshold value xi, wherein when xi is 0.5, the result of an outlier error is minimum, and the positioning accuracy of a target is highest;
when gamma is larger than or equal to xi, outputting an outlier sensor point set; otherwise, outputting the effective sensor point set.
7. The method of claim 6, wherein the adaptive K-means outlier de-constrained optimization method for fusion mesh LOF comprises: and calculating the square sum of the distances from each sensor node to the clustering center of the K-means algorithm, circularly calculating a plurality of times by taking the minimum square sum of the distances from each point in the region to the clustering center as a standard after the K-means algorithm is clustered, and counting the times exceeding the standard, namely the times of the sensor node clustering, and recording as M.
8. The method of claim 1, wherein the method comprises: the step 5 specifically comprises the following steps: the method comprises the steps of returning an outlier sensor node data set, carrying out subtraction operation on the outlier sensor node data set and an initial sensor node data set, screening out an effective sensor point set, inputting the effective sensor point set into a bistatic sonar system positioning model, and effectively positioning a target.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010850065.0A CN112130154A (en) | 2020-08-21 | 2020-08-21 | Self-adaptive K-means outlier de-constraint optimization method for fusion grid LOF |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010850065.0A CN112130154A (en) | 2020-08-21 | 2020-08-21 | Self-adaptive K-means outlier de-constraint optimization method for fusion grid LOF |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112130154A true CN112130154A (en) | 2020-12-25 |
Family
ID=73850978
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010850065.0A Pending CN112130154A (en) | 2020-08-21 | 2020-08-21 | Self-adaptive K-means outlier de-constraint optimization method for fusion grid LOF |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112130154A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116644373A (en) * | 2023-07-27 | 2023-08-25 | 深圳恒邦新创科技有限公司 | Automobile flow data analysis management system based on artificial intelligence |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2012253680A1 (en) * | 2011-05-06 | 2013-12-19 | Hadal, Inc. | Systems and methods for synthetic aperture sonar |
| CN104808207A (en) * | 2015-04-14 | 2015-07-29 | 西安理工大学 | Chaotic underwater sound positioning method |
| CN106707233A (en) * | 2017-03-03 | 2017-05-24 | 广东工业大学 | Multi-side positioning method and multi-side positioning device based on outlier detection |
| CN107870621A (en) * | 2016-10-25 | 2018-04-03 | 中国科学院沈阳自动化研究所 | Autonomous underwater robot collision prevention method in unknown complex sea-floor relief environment |
| CN108508444A (en) * | 2017-02-27 | 2018-09-07 | 国网山西省电力公司阳泉供电公司 | A kind of mining influence area power transmission line sag computational methods and system |
| CN109001737A (en) * | 2018-05-25 | 2018-12-14 | 哈尔滨工程大学 | A kind of Bi-static sonar system direct wave suppressing method |
| CN109446422A (en) * | 2018-10-24 | 2019-03-08 | 湖北大学 | A kind of service recommendation method based on the user filtering that peels off |
| CN110083665A (en) * | 2019-05-05 | 2019-08-02 | 贵州师范大学 | Data classification method based on the detection of improved local outlier factor |
| CN110362608A (en) * | 2019-06-11 | 2019-10-22 | 广东工业大学 | Energy consumption method for detecting abnormality based on rain flow method and local outlier factor |
| CN110636437A (en) * | 2019-09-05 | 2019-12-31 | 哈尔滨工程大学 | Fingerprint pair-based auxiliary positioning method |
| CN110824485A (en) * | 2019-09-25 | 2020-02-21 | 中国人民解放军军事科学院国防科技创新研究院 | Bistatic sonar underwater target feature extraction and classification method based on AUV |
| US20200211325A1 (en) * | 2018-12-31 | 2020-07-02 | Playtika Ltd. | System and method for outlier detection in gaming |
-
2020
- 2020-08-21 CN CN202010850065.0A patent/CN112130154A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2012253680A1 (en) * | 2011-05-06 | 2013-12-19 | Hadal, Inc. | Systems and methods for synthetic aperture sonar |
| CN104808207A (en) * | 2015-04-14 | 2015-07-29 | 西安理工大学 | Chaotic underwater sound positioning method |
| CN107870621A (en) * | 2016-10-25 | 2018-04-03 | 中国科学院沈阳自动化研究所 | Autonomous underwater robot collision prevention method in unknown complex sea-floor relief environment |
| CN108508444A (en) * | 2017-02-27 | 2018-09-07 | 国网山西省电力公司阳泉供电公司 | A kind of mining influence area power transmission line sag computational methods and system |
| CN106707233A (en) * | 2017-03-03 | 2017-05-24 | 广东工业大学 | Multi-side positioning method and multi-side positioning device based on outlier detection |
| CN109001737A (en) * | 2018-05-25 | 2018-12-14 | 哈尔滨工程大学 | A kind of Bi-static sonar system direct wave suppressing method |
| CN109446422A (en) * | 2018-10-24 | 2019-03-08 | 湖北大学 | A kind of service recommendation method based on the user filtering that peels off |
| US20200211325A1 (en) * | 2018-12-31 | 2020-07-02 | Playtika Ltd. | System and method for outlier detection in gaming |
| CN110083665A (en) * | 2019-05-05 | 2019-08-02 | 贵州师范大学 | Data classification method based on the detection of improved local outlier factor |
| CN110362608A (en) * | 2019-06-11 | 2019-10-22 | 广东工业大学 | Energy consumption method for detecting abnormality based on rain flow method and local outlier factor |
| CN110636437A (en) * | 2019-09-05 | 2019-12-31 | 哈尔滨工程大学 | Fingerprint pair-based auxiliary positioning method |
| CN110824485A (en) * | 2019-09-25 | 2020-02-21 | 中国人民解放军军事科学院国防科技创新研究院 | Bistatic sonar underwater target feature extraction and classification method based on AUV |
Non-Patent Citations (4)
| Title |
|---|
| TANG XINYU ET AL.: "Methods for Underwater Sonar Image Processing in Objection Detection", 《2017 INTERNATIONAL CONFERENCE ON COMPUTER SYSTEMS，ELECTRONICS AND CONTROL(ICCSEC)》 * |
| 张硕等: "基于网格LOF和自适应K-means的离群点检测算法", 《指挥信息系统与技术》 * |
| 汪晓亮: "双/多基地声呐定位技术研究", 《中国优秀硕士学位论文全文数据库（电子期刊）工程科技II辑》 * |
| 王丽娟等: "基于RSSI测距误差离群去约束的优化方法", 《河南农业大学学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116644373A (en) * | 2023-07-27 | 2023-08-25 | 深圳恒邦新创科技有限公司 | Automobile flow data analysis management system based on artificial intelligence |
| CN116644373B (en) * | 2023-07-27 | 2023-10-31 | 广东小途汽车科技有限公司 | Automobile flow data analysis management system based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhong et al. | Particle filtering approaches for multiple acoustic source detection and 2-D direction of arrival estimation using a single acoustic vector sensor | |
| CN106842165B (en) | Radar centralized asynchronous fusion method based on different distance angular resolutions | |
| CN111524084B (en) | Photon counting laser radar point cloud denoising method based on multi-peak Gaussian fitting | |
| Liu et al. | Wave height estimation from shipborne X-band nautical radar images | |
| CN111965632B (en) | Radar target detection method based on Riemann manifold dimensionality reduction | |
| CN110007299B (en) | Weak target detection tracking method based on mixed coordinate pseudo-spectrum technology | |
| WO2023116196A1 (en) | Aoa and tof joint estimation method and apparatus for indoor positioning, and storage medium | |
| JP6773606B2 (en) | Radar device and its radar signal processing method | |
| CN111337878B (en) | Sound source direct positioning method suitable for large-aperture horizontal linear array | |
| CN115061113B (en) | Target detection model training method and device for radar and storage medium | |
| KR101817011B1 (en) | Clutter reduction Method and Apparatus for active sonar based on clustering characteristics | |
| CN113267754A (en) | Three-dimensional grid-based terrain occlusion radar detection range calculation method | |
| CN110244260B (en) | Underwater target high-precision DOA estimation method based on acoustic energy flow vector compensation | |
| CN112130154A (en) | Self-adaptive K-means outlier de-constraint optimization method for fusion grid LOF | |
| CN115840205A (en) | Laser radar technology-based geosurface area metering method and system | |
| CN115220007A (en) | Radar point cloud data enhancement method aiming at attitude identification | |
| CN108983194B (en) | Target extraction and condensation method based on ground monitoring radar system | |
| Dubrovinskaya et al. | Anchorless underwater acoustic localization | |
| CN109239677B (en) | Environment self-adaptive constant false alarm rate detection threshold determination method | |
| KR102361816B1 (en) | Method for detecting target and readable medium | |
| CN112666528A (en) | Multi-station radar system interference identification method based on convolutional neural network | |
| CN116594018B (en) | Helicopter flight obstacle alarm system and method | |
| CN113640737B (en) | Two-dimensional power distribution-based small array element array high-resolution azimuth estimation method | |
| CN115453489B (en) | Indoor multipath discrimination method for millimeter wave radar | |
| CN111723797B (en) | Method and system for determining bounding box of three-dimensional target |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201225 |